Tacho issues

[quattro]

This appears to be what has been added - looks familiar somehow

I wonder of I can just add a relay to this

Edited October 14, 2017 by quattro

[lo-fi]

What vehicle is it, Quattro, and was it injected originally? 

I assume this used to be a single coil RV8 setup the tachometer was fed from? If that's the case, Quagmire's setup looks most likely to work. 

The circuit you've posted would work for something modern that's ECU driven as stock. The Lotus Elise, for example, has the tacho line held up at 5v by a resistor in the tacho, and is expecting the ECU to ground it to send a signal, which your circuit posted above is perfect for.

Older tacho circuits like I think you probably have work by detecting a voltage spike caused by the change in state of the coil primary, and are often set to trigger above a certain voltage for noise elimination. 

I'd try the relay coil using that circuit you've got - it'll probably work. 

[quattro]

Hi lo-fi

It's a P6 Rover and yes, it used to run from a single coil and carbs.

I have used Quagmire's circuit and it doesn't work, a Tachfix doesn't work, the four 1N4004s circuit works up to 1,300rpm. All of these are from the megasquirt manual and are for the single coil, high voltage tacho.

I have another tacho for a 4 cylinder P6, and the various fixes above, do the same with that.

The circuit I have posted in my last post is one that was added to the ECU when the output wire was fitted. This is inside the ECU.

I was remarking on the similarity of this to the circuit used by Quagmire.

Richard

[lo-fi]

Ah, a lovely machine! 

Do you know much about the tacho?  RVI or RVC? Apologies if I missed this info from the thread already. It'll need connecting and driving quite differently depending which type it is, so positively  ID'ing it will help. 

This thread suggests you may have the RVI type:

https://classicroverforum.net/index.php?threads/tachometer-question-again.38731

/Ian

[Quagmire]

Richard, it looks to me like you should be able to take a +12v feed, connect that to one leg of your relay coil, and connect the other leg to the wire that has been added to your MS, then you connect the old white/black wire that used to connect to the coil negative to the same place. 

 

 

[quattro]

Hi Ian

The tacho is the RVC type and worked with the dizzy and coil set up. I have a spare tacho, again an RVC type but for a 4 cylinder car. I have changed to this one to check and it does the same as the V8 0ne.

 

[quattro]

12 minutes ago, Quagmire said:

Richard, it looks to me like you should be able to take a +12v feed, connect that to one leg of your relay coil, and connect the other leg to the wire that has been added to your MS, then you connect the old white/black wire that used to connect to the coil negative to the same place.

So, use the circuit he added as part of the tacho fix circuit? I'll give it a go tomorrow

Richard

[lo-fi]

OK, thanks Richard. So what Quagmire suggests above *should* work. There are sometimes snubbing caps and suchlike inside relays to stop the coil making voltage spikes as we're actually looking for it to do - it really depends on the relay. 

I have played with the RVC type a little bit, and actually developed a circuit to drive an old Smiths 4cyl unit I found to fit in the dash of my V8 S3. I've got loads of boards left over, and it's only about a quids worth of components that goes into a circuit. A last resort, but you're more than welcome to add to your options list. 

I'm pretty sure Quagmire is leading you in the right direction though, so I won't muddy the waters. Might be worth trying a few different relays as they'll all have different characteristics is all I'd add. 

Good luck

[quattro]

I've added a relay and now have this



I took the switching gear out of the relay as it wasn't needed but this meant I couldn't listen to it to check it was working correctly, so I went out and bought a new one.

Started the car and the tacho sat at 750-800 rpm which was right but only for a few seconds and then it dropped, then started again, then stopped, and started again, etc.

I got out of the car and listened to the relay, and it buzzed, then stopped etc in line with the tacho, so if it buzzes, the tacho works.

I can only conclude that the signal switching the relay, from the ECU, is not switching it reliably.

I take it from the diagram, that the transistor is like a switch which is turned on and off by JS3 after going through a 1K resistor, and this then allows the current to pass through to earth, thereby firing the relay.

So any ideas why this intermittent feed is happening?

Thanks in advance

tachbounce.wmv

relay.wmv

[Bowie69]

WMV, how very retro

 

That is odd, you would expect it to continue buzzing.... 

 

I'm suspicious that either a setting in MT is wrong, or there is just an error in the board.. is the buzzing on/off regular?

[quattro]

27 minutes ago, Bowie69 said:

WMV, how very retro

I am a retro sort of chap

The buzzing on off is not regular at all, it's more of a sh... ermm fits and starts sort of thing, like it's trying to work but isn't quite there.

Richard

[Bowie69]

Tried wiggling things on the MS board while running? Or the main connector?

[quattro]

I'll give that a go tomorrow. I will also get the multi meter out and test the circuit along its entirety whilst it is running, hopefully find an intermittent voltage somewhere.

I'll be glad when I've had enough of this

[quattro]

Right tried wiggling everything, and the car engine died. So managed to start it again and took it for a short drive where it kept dying on me, so back to the workshop and tried wiggling everything again. Checked every plug, pin, wire, connection, nothing. So went and got some petrol next job after the tacho is the fuel gauge.

So is the 2N2222 up to the job?  - I am not good enough at soldering to try to remove the 2N2222 and replace it with the 2N5551 in the ECU   My soldering skills look like I have used an arc welder. Or, would it be sensible to cut both the red and orange wires in the ECU and solder them together, to bypass the circuit in there, and then have a wire direct from JS3. Then run that through my circuit.



Also, I have googled the output system and found that the settings may be wrong .

I have changed the figures in the boxes from 7, 1, and 0 to 5, 2, and 55, which should apparently start the feed from JS3 at 500rpm, stop it if it gets down to 300rpm and give it a higher limit of 5,500rpm. This made no difference at all so does anyone know if these figures are correct?

Thanks in advance

Richard
 

 

[lo-fi]

Whatever you do, don't plug JS3 directly into a coil! You'll probably fry the processor, or at very least that output. The 2n2222a is there to do all the heavy lifting and should be more than up to the job unless you're using a whopping great coil. It's well chosen for a switching application and that circuit is fine, albeit lacking flyback protection, but we'll get to that later. 

The MS settings look fine, I think. 

I'd remove the mechanism from the relay as you did before, it may be moving into a position where it resonates and does funny things with the field.. I'm speculating there, but relays do odd things when asked to switch quickly. At a certain frequency they tend to "stick". It would be better to verify the electronic part is working with an LED and resistor, rather than listening for the relay. They're not designed to switch fast! 

I think what you're struggling with is getting a consistent voltage spike high enough to trigger that tacho - it seems to want quite a lot - not a configuration or electronic issue: I'd put money on that. Getting closer, though

[quattro]

I would be using JS3 to run through this :-

So, not directly to the coil.

The firing of the relay is intermittent, you can hear it with the switch gear in there. It makes no difference to what the tacho does, whether I use the relay with the switch still in there, or the one I have stripped out. I have three relays in all and they all do the same. When the relay is switching on and off (buzzing), the tacho works correctly, so I am thinking the circuitry somewhere isn't up to firing the relay.

I will try to get a voltage at both ends of the relay when the engine is running, and see what is happening there, but am short of time at present (I'm supposed to be working )

Richard

 

 

[quattro]

Out with the multi meter - right with this circuit the relay buzzes intermittently. With the engine running, the voltages are +ve side of coil (85)- 14.3V, -ve (86) - jumped from 0.5V to 9V and all points in between. Voltage at JS3 jumped around from 1V to 4.5V.

I disconnected the orange wire, and then connected a patch wire to it, then clipped the other end directly to JS3 with a crocodile clip. This basically bypassed the circuit in the ECU without me having to remove/cut anything . I then connected this to my circuit board like this

Now, the voltages are (85) - 14.3V, (86) - 5.4V (pretty much constant), JS3 4.5V, base of 2N5551 0.7V, resistor in 4.5V, resistor out 0.7V. All of these voltages are fairly constant.

The tacho does now point in the right general direction but still waves around a fair bit. It's better, but not yet right, and not accurate enough to use. There is not enough voltage to fire the relay, so no buzzing at all. All the above readings are at tickover (750-800rpm).

So, more from intuition than electronic knowledge, I would say there's something amiss within the added circuit in the ECU - although any opinions would be greatly received

[garrycol]

Wow - this is getting bigger than the Ben Hur production - 60 odd posts over two forums and still resolution.

[Bowie69]

I can remember a mate's old P6 used to have a pretty bouncy/wayward tach, have you ever had it working properly before MS was installed?

[quattro]

18 minutes ago, Bowie69 said:

I can remember a mate's old P6 used to have a pretty bouncy/wayward tach, have you ever had it working properly before MS was installed?

Hi Bowie. It worked perfectly prior to the installation of the EFI. Also, I do have another tacho, albeit from a 4 cylinder P6 (still RVC), and when I fitted that it bounced around like the 8 cylinder one.

I will try that again now I have got somewhere though, I am open minded about the cars original wiring.

Richard

[quattro]

4 hours ago, garrycol said:

Wow - this is getting bigger than the Ben Hur production - 60 odd posts over two forums and still resolution.

I was advised to post on other forums Garry, and Dave has been very helpful. The third forum haven't answered yet so I'll just stick with the two for now.

[garrycol]

Yes I agree with you about Dave - he is very helpful and not backward in giving his time to help others.  Following the discussion on the V8 forum - what is the 3rd?  The MS Forum?

Garry

 

[quattro]

Megasquirt support forum (MSEXTRA).

It does look like there is something wrong with the added circuitry in the ECU, as the voltage is bouncing around inside it.

My circuit is not opening the transistor fully as I'm only getting 5.4V at the -ve side of the relay? I'm going to try and whack it with a hammer later

Richard

 

[Quagmire]

Seems like you are making progress! 

[lo-fi]

Hi Richard

Something not quite adding up here. What voltage are you measuring and how? Are you trying to measure AC or DC at the coil>collector joint? Both are bound to give  "odd"  readings on a standard multimeter, and absolute values won't necessarily be that helpful. 

I'm a bit confused about your circuits too, when you were explaining about  swapping the wire connection... That  2n2222, is that as well as another transistor outside the case? Or were you trying another transistor instead of that? 

Oh, and one last thing... (someone will get the reference, I'm sure ) does the relay cut in and out as revs rise and fall slightly? Does the tacho continue to work above idle too? 

I think you're pretty close